Textkonventionen am Bundesverfassungsgericht auf der Spur

The linguistic form of decisions of the German Federal Constitutional Court remains largely unexplored. In this dissertation, the author undertakes a systematic empirical analysis that combines close readings of individual texts with large-scale computational methods. The foundation is a corpus of around 3,400 rulings from the Official Digest and 6,900 additional decisions of the German Federal Constitutional Court. Special attention is given to text segments such as rubric, tenor, reasoning, statement of facts, admissibility, merits, and the so-called "standards section". These are annotated either automatically or manually and analysed with respect to their frequency, length, and linguistic properties. The findings show that the degree of conventionalisation varies according to type of decision, outcome, and judicial body, with official rulings being more standardised than chamber decisions. Norm control procedures, for example, produce longer and more complex structures. The study also identifies diachronic developments: texts become longer, more uniform, and more consistently organised, with introductory summary sentences increasingly established. Initial peculiarities-such as missing text segments or unusual subdivisions-disappear over time. By applying digital methods to constitutional jurisprudence, the work uncovers patterns invisible in single-case studies, confirming some scholarly intuitions while challenging others. It thereby opens new avenues for both linguistic and legal research and exemplifies the effective intersection of law and the digital humanities.